A Survey to Investigate Student Drinking Norms at Foxfield Neil Guha, Ellen J. Bass, Senior Member, IEEE, Susan E. Bruce Abstract— Drinking plays a role in the social activities of University of Virginia (UVA) students, including the annual Foxfield Horse Races. However, no data gauge the levels of binge, or high-risk, drinking at Foxfield. A web-based survey has been developed to begin to measure the drinking norms at Foxfield with the hopes of using the data to create a safer environment at future events. Using a limited sample of 78 students, the preliminary data show that students drink more at Foxfield than each day of a normal week. Students also report drinking more and longer at Foxfield than during Halloween and football games (events known to be large social events at UVA). 66% of respondents also reported at least one negative consequence as a result of another’s drinking and 68% reported at least one negative consequence as a result of his own drinking. This work helps to lay the groundwork for a large sample data collection after the 2006 running of the Foxfield Races on April 29th.
B
I. INTRODUCTION
inge drinking, or high-risk drinking, is generally defined as consumption of alcohol that causes negative consequences to oneself and others. It has been quantified as 5 or more drinks in two hours for a man and 4 or more drinks in two hours for a woman [1]. Estimated Blood Alcohol Concentration (eBAC) is a more exacting method used to quantify the amount of alcohol consumed by an individual and is calculated differently for men and women: • Women: eBAC= (number of drinks)/2 * 7.5/(weight in pounds) – 0.16 * (hours spent drinking) • Men: eBAC= (number of drinks)/2 * 9/(weight in pounds) – 0.16 * (hours spent drinking) According to the National Institute on Alcohol Abuse and Alcoholism, binge drinking contributes to at least 1,400 university student deaths and 500,000 injuries per year [2]. Due to the large negative outcomes with binge drinking at universities, researchers are interested in understanding this phenomenon. For example, the Harvard School of Public Health conducted a survey in 2001 of 119 four-year colleges to track heavy alcohol use. The study found that Manuscript received April 14, 2006. Neil Guha is an undergraduate student at the University of Virginia (e-mail:
[email protected],
[email protected]). Ellen J. Bass is an Assistant Professor in the Department of Systems and Information Engineering at the University of Virginia, P.O. Box 400747, 151 Engineer’s Way, Charlottesville, VA 22904 USA (e-mail:
[email protected]). Susan E. Bruce is the Director of the Center for Alcohol and Substance Education at the University of Virginia, 2400 Old Ivy Road, Suite C, Charlottesville, VA 22904 USA (e-mail:
[email protected]).
approximately two out of every five college students reported binge drinking [3]. This confirmed previous studies at the same colleges in 1993, 1997 and 1999. Binge-drinkers were also found to be 21 times more likely than non-binge drinkers to have missed class, to have been injured or hurt, or to have driven a car intoxicated. At colleges with high binge drinking rates, 71% had sleep or study interruptions caused by drinkers and 57% had to take care of an intoxicated student [3]. The Department of Psychiatric Medicine and the Center for Alcohol and Substance Education (CASE) conduct a yearly Health Behavior Study (HBS) to gauge controlled substance norms at the University of Virginia [4]. Both the results from HBS and prior research have shown that high levels of binge drinking contribute to unwanted negative effects from both one’s own drinking and someone else’s drinking. Although fatality is an extreme consequence of alcohol consumption, binge drinking is also associated with numerous other negative consequences. For example, in 2003, 46% of UVA students said they assumed a caretaking role and 44.1% had their sleep disrupted as a result of someone else’s drinking [4]. Additionally, the Office of Health Promotion at UVA conducts a yearly Health Promotion Survey (HPS) which focuses solely on alcohol use and protective behaviors [5]. The study has shown an increase in median drinks per week for all UVA undergraduates from 2 in 2001 to 3 in 2003 (Figure 1). 5 4.5 4 3.5 3 2.5 2 1.5 1 0.5 0
Men Women Total
2001
2002
2003
Fig. 1. Median weekly drinks for UVA Undergraduates (2001-2003) [5].
At the University of Virginia, drinking plays a role in students’ social activities, including the Foxfield Races. Thousands attend the annual event held in April, many of whom are students at the University of Virginia. Students who attend do so individually or with an organization, which may purchase a plot. Attendees bring food and both
alcoholic and non-alcoholic beverages in order to picnic from approximately 10am to 4pm. Foxfield has taken several preventative measures to help curb binge-drinking and its consequences such as incentives for sober drivers, ability to leave vehicles at the venue overnight, and increasing law enforcement (with plans to increase police presence for the 2006 races). Foxfield also has a first-aid tent located by the entrance to the plots for anyone who is seeking medical attention. There are no recorded data on drinking, preventative measures and their effects at the Foxfield Races. The purpose of this project is to conduct a web-based survey that collects data on drinking norms at Foxfield. CASE is interested in seeing if students drink more at Foxfield than during the course of a week and during other eventful weekends at UVA. It is also useful to examine the extent to which students take preventative measures such as taking advantage of the first-aid tent and consuming foods and nonalcoholic beverages. CASE is interested in finding out if some preventative policies, such as the Savvy Fox program which gives free food to sober drivers, may actually increase the levels of high-risk drinking among non-drivers. CASE is also interested in understanding the negative consequences students report as a result of one’s own or someone else’s drinking and if affiliation with organizations impact drinking-related behaviors. To accomplish these goals, the survey provides data to test the following specific hypotheses: 1. Student drinking at Foxfield is greater than student drinking during the course of a normal week at UVA as measured by average number of drinks consumed and eBAC 2. Student drinking at Foxfield is greater than student drinking during eventful weekends at UVA (Halloween) as measured by average number of drinks and eBAC 3. Student drinking at Foxfield is greater than student drinking at a UVA home football game as measured by average number of drinks consumed and eBAC 4. Drunk driving prevention measures cause an increase in alcohol consumption among non-drivers as measured by average number of drinks consumed and eBAC 5. Members of Greek-affiliated organizations drink more than non-Greek organizations at Foxfield as measured by average number of drinks consumed and eBAC 6. An increase in alcohol consumption increases the risk for negative consequences both from personal drinking and secondhand effects from someone else’s drinking 7. Attending Foxfield more than once reduces eBAC and the average number of drinks consumed by an individual
This paper focuses on designing and testing a web-based survey on drinking norms at Foxfield by conducting a preliminary data collection and analysis. A larger-scale sample will be collected and analyzed after the 2006 running of the Foxfield Races on April 29th. II. METHODS A. Institutional Review Board Approval was granted by the Institutional Review Board for Social and Behavioral Sciences (IRB-SBS) before survey data were collected. B. Sample For purposes of testing the survey, a “Beta Test” population was chosen. This population was pooled from three sources: • EDHS 224, Substance Abuse in Society • SYS 334, Systems Evaluation • Sigma Nu Fraternity EDHS 224 is a course on substance education taught in the Curry School of Education. It is a class of predominantly fourth-years although a few underclassmen are enrolled. SYS 334 is a third-year course required for all Systems Engineering majors. Sigma Nu is a national fraternity and is a member of the Inter-Fraternity Council at UVA. The potential respondents in the two courses were contacted via the class toolkit e-mail and Sigma Nu was contacted via the fraternity group e-mail list. They were informed that all data collected would be stripped of identifying information and be anonymous. Furthermore, all potential respondents were informed that their participation is voluntary. EDHS 224 has an enrollment of 152 with 27 respondents (response rate of 17.76%). Sigma Nu has a brotherhood of 70 with 23 respondents (response rate of 32.86%). SYS 334 has an enrollment of 84 with 28 respondents (response rate of 33.33%). Thus the total sample size for the Beta Test was N = 78. Table I identifies demographic data for the respondents according to year at UVA and gender. TABLE I BREAKDOWN OF BETA TEST SAMPLE Type of # of % of Sample Group Respondents (N = 78) 1st Years
8
10.25%
2nd Years
8
10.25%
3rd Years
33
42.30%
4th Years Graduate Students
28
35.89%
1
1.20%
Males
47
60.26%
Females
31
39.74%
C. Survey Instrument The 24-question survey instrument was developed in a collaboration by the authors and other CASE professionals. It was iteratively improved following a pilot test by a 20person sample. It is divided into three main sections: • Background Information • UVA-Related Questions • Foxfield-Related Questions Each of the 24 questions either helps test one of the 7 hypotheses or provides CASE data to compare with prior studies. The first section includes 5 questions: year at the university, gender, age, weight, and extra-curricular activities. Gender and weight help calculate eBAC and the activities involved help test Hypothesis #5. Although this hypothesis essentially measures drinking norms of Greeks vs. non-Greeks, the question supports comparisons between other groups and organizations (which will be helpful in the larger sample study). The second section contains 6 questions attempting to measure drinking norms at non-Foxfield events. Questions #6-11 include average number of drinks consumed and hours spent drinking each of day of the week, average number of drinks consumed and hours spent drinking at UVA home football games, and average number of drinks consumed and hours spent drinking during Halloween. These questions target Hypotheses #1-3, comparing drinking norms at Foxfield to each day of a normal week, to the day of a home football game and Halloween. Football game-day was chosen as a measure of comparison because it is a staple University event. Halloween was chosen as a measure of comparison because it is a social non-University related event, similar to Foxfield. The third section, Questions #12-24 gather data about drinking norms at the Foxfield Races. Question #12 asks the respondents whether or not they have attended Foxfield. Questions #13-24 only apply to respondents who have attended Foxfield. Question #13 asks the respondents how many times they have attended Foxfield. This targets Hypothesis #7 to analyze norms across those who have attended the yearly event more than once. Questions #14-15 asks the average number of drinks and hours spent drinking at the races. These data are required to test each of the 7 hypotheses. Questions #16-17 are related to transportation methods and gauging levels of drunk driving. Question #16 asks the respondents to select their method of transportation to and from Foxfield. The options provided are: 1) I was a designated sober driver to and from Foxfield; 2) I rode the CTS bus; 3) I had a sober friend drive me to and from Foxfield; 4) I drove to Foxfield but left my car there overnight and found another transportation source back; 5) I drove to Foxfield and back because I did not think I drank enough to impact my driving; 6) I drove to Foxfield and back even though I had too much to drink;
7) My group chartered a bus; 8) Other Please Specify: This question contributes to Hypothesis #4 which tests whether policies instituted to help prevent drunk driving may cause an increase in binge-drinking among non-drivers. Question #17 simply asks if an individual may be inclined to drink more if he knew he were being driven to and from Foxfield by a sober driver. Questions #18-19 ask the respondent to check all the negative consequences that apply as a result of another’s drinking and his own drinking. These data target Hypothesis #6 which assumes that an increase in alcohol consumption yields an increase in negative consequences. The options provided for consequences as a result of another’s drinking are: 1) Placed me in a caretaking role (helping someone who is sick, helping someone get home, etc.); 2) Damaged my personal property (car, clothing, etc.); 3) I was a passenger in a vehicle driven by a drunk driver; 4) I experienced physical pushing, shoving or hitting; 5) Jeopardized a relationship (caused a verbal argument); 6) I experienced an unwanted sexual advance; 7) Disrupted my study time; 8) Disrupted my sleep; 9) Prevented me from enjoying the horse races; 10) Other Please Specify: The options provided for consequences as a result of personal drinking are: 1) Had a hangover; 2) Was nauseous or vomited; 3) Was taken to the ER (UVA or other); 4) Got into a physical fight; 5) Took advantage of someone sexually; 6) Experienced unwanted sexual advance or sexual assault; 7) Drove a car under the influence; 8) Was arrested for DUI/DWI 9) Been in trouble with police for something OTHER THAN DUI; 10) Urinated in public; 11) Blacked out (had memory loss); 12) Other Please Specify: Question #20 simply asks if the respondent knows where the first aid tent is. Questions #21-22 asks the respondent what foods were consumed before attending Foxfield and while at Foxfield. The options for both these questions are the same: 1) Water; 2) Juice (without alcohol mixed); 3) Soda (without alcohol mixed); 4) Fruits; 5) Vegetables; 6) Bread product; 7) Dairy product; 8) Meat; 9) Other Please Specify:
D. Data Analysis The data from the Beta Test were analyzed using SPSS Version 13.0. If a respondent skipped an individual question or provided an invalid response, that response was removed from any tests associated with the particular question. Descriptive statistics were derived for the survey questions and t-tests were run based on the original hypotheses. All data that reported never having attended Foxfield were excluded from all Foxfield-related analyses, but were still used for calculating means related to Questions #1-11. III. RESULTS All results are report with an alpha level of 0.05. The first analysis compared average number of drinks consumed throughout the course of the week vs. average number of drinks at Foxfield (Hypothesis #1). Table II row 2 presents the baseline Foxfield data for number of drinks consumed and rows 3-9 show the two-sided t-tests that compared each day of the week to the baseline. TABLE II DESCRIPTIVE STATISTICS AND T-TEST RESULTS FOR DRINKS CONSUMED DURING DAYS OF THE WEEK, HALLOWEEN AND FOOTBALL GAMES VS. FOXFIELD Event n Mean Std TPDF Dev Ratio Value Foxfield 54 11.61 6.65 Monday Tuesday
74 75
0.50 1.80
1.31 2.63
12.11
